A Probe Shift Is Not a Fairness Fix: The Limits of Representation Steering in Speech Models
This work probes every encoder layer for metadata-derived sex/gender, age, and native/accent labels; construct centroid and probe-derived directions; inject them at selected layers; and compare downstream probe trajectories with matched WER changes.
Nicolas Bourrel, Abderrahmane Issam, Gerasimos Spanakis
· 0 citations